Essays on equilibrium unemployment dynamics

Abstract

This thesis is a collection of three essays in which the behaviour of unemployment is studied in different dynamic environments. Throughout, unemployment is understood to be involuntary, arising due to the uncoordinated nature of trade in the labour market as viewed from the perspective of the Diamond-Mortensen-Pissarides equilibrium matching model. It goes without saying that the fundamental motivation for pursuing this line of research is provided by the untold consequences, both human and economic, of otherwise capable people remaining involuntarily idle. An attempt, therefore, is made to contribute to the understanding of how various aspects of macroeconomic policy can influence unemployment outcomes. The approach maintained throughout is to combine general equilibrium modelling with simulation techniques in order to provide not only qualitative inferences but also quantitative descriptions of equilibrium dynamics. The dynamic environments considered cover both the business cycle (the first two chapters) and the life cycle (the third chapter). In the first chapter, Structural Tax Reform and the Cyclical Behaviour of the Labour Market, we build a real business cycle model with frictional unemployment and distortionary tax rates which are increasing in individual taxable labour income. The cyclical aspects of tax reform that are addressed in this chapter are distinct from the stationary state distributional issues that have garnered most of the attention in the existing literature on structural tax reform. Estimating the tax code parameters from federal income tax return data for the U.S., we find that a reduction in the progressivity of the tax system is associated with a significant increase in the volatility of hours per worker. The intuition is simply that the greater the extent to which marginal tax rates fluctuate in response to shocks, the smaller the incentive to adjust working hours. But in a frictional labour market in which it is costly for forms to issue vacancies, the behaviour of hours - i.e. intensive adjustment, or adjustment in the intensive margin - is a determining factor of job creation - i.e. extensive adjustment. We then explain how the dynamic behaviour of hours along the adjustment path to an aggregate productivity shock generates o¤setting incentives for job creation, with the result that tax reform has little impact on unemployment fluctuations. The welfare cost of the business cycle is also computed under different tax regimes. It is found that although business cycles are more costly under a flat tax, the overall welfare implications are quantitatively negligible regardless of the tax system. Having described the effects of the tax system on equilibrium dynamics when perturbed by a productivity disturbance, we then consider business cycle adjustment to an aggregate demand shock in the form of fiscal stimulus. In light of recent fiscal developments in the U.S. and Europe, the ability of expansionary fiscal policy to stimulate output has gained renewed interest in the business cycle literature. We contribute to the analysis by assessing whether the efficacy of government expenditure in reducing unemployment depends on the structure of the tax system. It is demonstrated that a less progressive tax policy increases the ability of expansionary fiscal policy to stimulate output due to a larger response in hours, but this comes at the cost of a smaller unemployment multiplier. Tax reform therefore causes a compositional shift in labour market adjustment in response to aggregate demand shocks, with relatively more adjustment occurring in the intensive margin and less adjustment in the extensive margin the flatter the tax schedule is. The reason why this compositional shift occurs for a demand shock but not a supply shock is that the adjustment path of hours is qualitatively dependent on the type of disturbance. In particular, we describe how equilibrium undershooting in hours occurs only in response to an aggregate productivity (supply) shock, whereas the negative wealth effects arising from increased government expenditure exert sustained upward pressure on hours along the entire adjustment path, thus providing a significant incentive for firms to substitute away from job creation. The second chapter, Monetary Policy and Job Creation in a New Keynesian Model, is motivated by the work of Cooley and Quadrini (1999) and Krause and Lubik (2007). These studies indicate that a typical monetary business cycle model with frictional unemployment and endogenous job destruction tends to encounter difficulty in generating a rise in job creation in response to expansionary monetary policy, rendering the model inconsistent with the downward sloping Beveridge curve that appears in the data and implying only a limited policy role for inflationary job creation. Matching frictions in the labour market congest the job creation process so that firms tend to skew adjustment to shocks towards the job destruction margin. In recognition of the assertion put forth but unpursued by Cooley and Quadrini (1999) that fluctuations in the size of the labour force may ease labour market congestion and therefore amplify cyclical job creation, in Chapter II we extend a New Keynesian model with unemployment to feature an endogenous labour market participation decision. However, a baseline model with a standard degree of risk aversion tends to exhibit countercyclical labour force participation, which is inconsistent with the data. In order to address this issue, we propose the notion of labour market participation as a social consideration, which we demonstrate to be capable of generating procyclical participation incentives. The basic idea is that agents will tend not to exit the labour force during booms in order to "keep up with the Joneses". We then find that plausible fluctuations in the size of the labour force do not exert a quantitatively significant effect on job creation. In light of this result, we search for alternative mechanisms which may overturn the conclusion that inflationary policy is incapable of incentivising job creation. The approach taken involves switching focus to the characteristics of aggregate demand dynamics along the adjustment path to a monetary shock. It is well known that standard New Keynesian models fail to deliver the gradual, hump-shaped adjustment path to monetary policy shocks that is observed in the data. We argue that if aggregate demand experiences a persistent increase in response to a monetary shock instead of peaking on impact, the incentive for firms to create jobs becomes amplified. The intuition is that, since the job creation decision is forward-looking due to the presence of matching frictions, aggregate demand must rise persistently even after the shock takes place so that firms anticipate a further increase in aggregate demand in order for the time consuming process of issuing a vacancy to be justified. To demonstrate this, it is shown that, by altering the dynamics of aggregate demand, time-inseparability in the utility function can significantly improve the ability of expansionary monetary policy to increase job creation, allowing the model to generate a downward sloping Beveridge curve conditional on monetary shocks. In the appendix to Chapter II, we lend further credence to this hypothesis by describing how the manner in which monetary policy it- self is specified may give rise to hump-shaped adjustment dynamics and, consequently, amplify inflationary job creation. Finally, in Chapter III on Equilibrium Matching and Age Discrimination Policy, we abstract from business cycle issues and concentrate instead on the life cycle. Federal legislation prohibiting the discrimination of workers on the basis of age has been in place in the United States since the 1967 Age Discrimination in Employment Act. Yet empirical studies which aim to estimate the employment effects of such legislation have yielded inconclusive results. We approach the issue from a different perspective by deriving quantitative predictions of equilibrium unemployment theory to investigate how age anti-discrimination legislation impacts labour market performance. We do not seek to measure the impact of a particular episode of legislative reform, but aim to quantitatively explore the general equilibrium consequences of restricting the ability of employers to hire on the basis of age. The main conclusion is that an equilibrium matching model of the life cycle predicts a moderately positive effect on the employment rate of workers very close to retirement, but the overall impact of age discrimination policy on the life cycle pattern of employment is quantitatively small. This occurs because in a frictional matching equilibrium, the incentive to discriminate against workers closer to retirement is offset by labour market congestion, preventing the demand for older workers from falling excessively even when it is possible to discriminate on the basis of age. If the demand for workers of a particular age were to fall sharply, the rate at which a given vacancy is matched with a worker of that age cohort increases, allowing firms to capitalise on quick vacancy transition rates stemming from weak competition in hiring. The model thus suggests that the extent to which the effect of a finite horizon is offset by labour market congestion is quantitatively significant, implying a modest role for age discrimination policy in shaping the life cycle pro le of employment. Welfare issues are also addressed. In particular, we demonstrate that an age- dependent inefficiency arises in the labour market participation decision of finitely-lived agents when firms are not able to discriminate in the hiring process on the basis of age. The intuition is that older workers, for whom only a relatively short productive time horizon remains, do not internalise the negative effect that their participation decision has on the age distribution of the aggregate pool of searchers. However, the size of this externality is quantitatively negligible. It is emphasised that the analysis in Chapter III concerns solely the macroeconomic implications of age discrimination policy as separate and distinct from the issue of fairness which naturally arises in this context. Although the economic impact of age discrimination policy on employment is argued to be quite small, this does not of course imply that such policy does not have significant merit with respect to its assurance of the basic tenet of equal opportunity for all.